AI: What AI Means For The Future Of American Education

0:00This message comes from NPR sponsor Raymond James, a firm where wealth management, banking and capital markets solutions start with the connections between people. That's the power of personal. Disclosures at RaymondJames.com.

0:22Artificial intelligence is rapidly changing how we work, live, and engage with the world, whether that's using chat GPT to do our taxes or spending hours sifting through AI-generated images on social media. AI is also reshaping our education system, and some of you have feelings about that. My name is Julia Janik. I'm in Richmond, Virginia. I'm a fourth grade teacher in public schools, weighing in on the use of AI as a teacher. And I think it's excellent. We've been encouraged to do it using certain platforms and always double checking. But when I upload a standard into it and tell it what I want, it instantly gives me at least a beginning framework rather than restarting with a blank slate.

1:08It frees up time, and time is the one thing that teachers need most. Hi, this is Carly Buxton. I teach business communication at the undergraduate business school at the University of Richmond in Virginia. And I feel like this is a moment of existential crisis for educators like me because we are preparing students for a world in which they are likely to rely on AI as central to their working lives every day. But we are expected to police the ways that they are using it in the classroom or for assignments. So it really, it feels hypocritical to me, and it's such a struggle for us. Roughly half of U.S.

1:43students said they used AI in some way for school. That's according to a 2025 survey from the RAND Corporation. But some students are not on board. These days, I don't really touch it. I don't think it's very conducive to the building of cognitive power and your own ability to think and study. That's high school senior David McCann Lienaris of Pickerington High School in North Ohio. We'll hear more from him later. But first, how are students and teachers around the country using AI? And what does it mean for the future of school and how we define what it's actually for? I'm Jen White. You're listening to the 1A Podcast.

2:21That and more after this short break. Stay with us.

2:52itself. Visit betterhelp.com slash reviews. This message comes from Progressive Insurance. Insurance isn't one size fits all. That's why drivers have enjoyed Progressive's Name Your Price tool for years now. With the Name Your Price tool, you tell them what you want to pay and they'll show you options that fit your budget. So whether you're picking out your first policy or just looking for something that works better for you and your family, they make it easy to See your options. Visit Progressive.com. Progressive Casualty Insurance Company and Affiliates. Price and coverage match limited by state law.

3:26Let's meet our guests. Joining us is Leigh Gaines. She's reporting on education and how artificial intelligence is reshaping classrooms for NPR. Leigh, it's great to have you. It's lovely to be here. Also with us is Justin Reich. He's the director of MIT's Teaching Systems Lab. He's also the author of Failure to Disrupt, Why Technology alone can't transform education. Justin, welcome to the program. Thanks for having me. And Jeffrey Riley. He's the executive director of Day of AI. That's a nonprofit that offers free curricula designed to introduce K-12 students to AI. He's also the former Massachusetts Commissioner of Elementary and Secondary Education.

4:06Jeff, welcome to the program. Great to be here. Lee, let's just start with how we're seeing AI show up in classrooms right now. What are you finding in your reporting? Sure. So AI is ubiquitous in schools, both among teachers and students. My team and I at NPR, we partnered with Ipsos recently to do a survey of how K-12 teachers are using AI. And we found that 62 % of the respondents said that they've used it for work. Now, the interesting thing about that is that they don't appear to be using it with their students in the classroom. A lot of teachers right now are using it to help them with administrative tasks, sending emails, creating lesson plans, diversifying their lesson plans to meet students where they're at, that sort of thing.

4:55But on the other hand, we know that students, there's plenty of surveys out there that show that students are using it on their own to help them with their schoolwork. And then there's a smaller portion of teachers out there that are experimenting with students in the classroom using AI. A lot of those tools are specifically designed for education. It's not that common for a teacher to just bring chat GPT in the room. Instead, they're using platforms like Magic School or School AI, these companies that have designed education-specific AI tools for classrooms. Justin, how much does this usage vary, not just state by state, but school by school?

5:38The usage of these technologies is extremely varied across schools. There are places that are highly concerned about the introduction of these tools and the risks that they have for students bypassing cognition and not doing the work that they were doing before. And so there are lots of places where folks are going back to blue books, going back to pencil and paper, policing activities extensively. And there are other schools that say, you know, this is a major part of young people's future and we need to figure out ways of integrating it across our classes, not just in the computer science curriculum, but where does it fit in English?

6:16Where does it fit in history? Where does it fit in science? Where does it fit in the arts? So you'll find a little bit of everything across the country's 130 ,000 schools. And Lee, I just want to be really clear that what we're talking about here are three main buckets for how AI is showing up in school. You have educators who are using it for administrative purposes, sometimes for lesson plan development. There's AI literacy where learning AI is the focus. And then there are students who are using tools like ChatGPT on their own without the direction or maybe even knowledge of the school. Does that cover everything?

6:52I think so. And then there are teachers, Justin mentioned this, that are changing how they approach the work they do in the classroom in response to AI, not necessarily incorporating AI into their own workflows or into the classroom, but what they're doing instead is trying to redesign how classrooms operate to guard them against student AI use. Jeff, one of your organization's primary goals is to educate students and teachers on AI. And we'll talk more about why you think AI literacy is important a bit later, but what kinds of lessons and curricula do you help schools develop to teach AI? So we have curriculum that goes from K through 12.

7:33We make no assumptions that people have to start in kindergarten. But fundamentally, when I speak publicly about this, we talk about historically the three R's in education, reading, writing, arithmetic. That one's a bit of a stretch. And now I say it's now the four R's, reading, writing, arithmetic. And if you allow the stretch to continue, artificial intelligence skills, because we think kids are going to need to know this, but not at the expense of having a knowledge base first, AI literacy second, and then going to tools later. Well, Justin, you spoke to 120 teachers and students across the country to find out how they feel about AI in the classroom.

8:14And that includes Sarah Fawes, who's a high school English teacher in San Francisco. This is part of the concern is like, then you just become a paranoid person. You distrust every single thing your students give you. You start to kind of move through the world with this sense of like, everybody's cheating. Everybody's cheating. Nobody's doing the work. I can't trust anybody. And that is a terrible, terrible place to be. I really love my students. And I really want to believe that they're good and ethical people who care about their learning and their process. And then I hear myself being like, I'm pretty sure everybody is cheating.

8:50Now, Justin, before you asked teachers and students about AI, you made sure to first ask how their school year was going. Why was that important to understand the larger academic context before asking about AI? So these interviews are from a project called The Homework Machine, which we released as a seven-part podcast series at teachlabpodcast.com. And schools across the United States are just unbelievably varied. Sarah Falls teaches in the San Francisco area in a school where her building doesn't have any hot water. There are other people who teach in brand new school buildings with every resource that you could possibly imagine.

9:30If you're in one of those schools where your students come fed and their parents are employed and have health insurance and all those kinds of things, then AI might really be one of the most signature challenges, particularly the surge of students who are using large language models to do all of their homework and good bits of their classwork. Now, if you're in one of those schools where students are vulnerable and their parents to deportation, where kids are homeless, where they don't have enough to eat, then the arrival of large language models might be the 17th or 19th thing on your list of most important concerns.

10:08So the context, you know, schools in the United States are incredibly unequal. And that inequality, they're unequal in the resources that the communities have, unequal in the resources that the schools have. And that inequality really shapes how schools receive these new technologies. So, Leanne, you're reporting how big of an issue is academic integrity, like we heard from Sarah in that clip, when it comes to how students are using AI. How are schools even defining academic integrity? Yeah, it's a huge issue. Another thing that we found in our survey with Ipsos was that a lot of teachers are having a much harder time now distinguishing how much their students are actually learning because of the rise of AI.

10:51And I would also say that I think the academic integrity issue is related to an issue of trust. That's another thing that we saw and something that tape that we just heard really spoke to, which is that because of AI, teachers are less trustful of their students. And I would say that also works in the opposite direction, too. So a good example of this is I spoke to a teacher in California who used to send her students out to do beach cleanups, habitat restoration. She's a biology teacher. And she would give them extra credit for doing that work. And then her son showed her basically how easy it would be to fake a photo using AI, showing that they were at a registration table.

11:39And so she had to stop doing that. And that was sad. Coming up after a quick break, we hear from one student who's been thinking about AI a lot and how we should approach it in classrooms.

11:54This week on the NPR Politics Podcast, another progressive Democrat edges out an establishment candidate. Abdul El-Sayed won Michigan's Senate primary, campaigning on affordability, health care and ending U.S. military aid to Israel. Join us as we break down his victory and what it could signal for November's midterms on the NPR Politics Podcast. Listen on the NPR app or wherever you get podcasts. Zendaya and Tom Holland are back in Spider-Man Brand New Day. And the movie has a more grounded tone than we've seen from the franchise. But does it strike the right balance of fun and darkness? We'll talk about it on NPR's Pop Culture Happy Hour.

12:31Listen via the NPR app or wherever you get your podcasts. Iran, Lebanon, Israel, and Gaza. With conflict unfolding in so many places, firsthand reporting has never mattered more to listeners who want to know the truth about what's happening around the world. NPR Plus supporters provide the crucial support our newsroom needs to cover the world's most important stories. You can learn more about NPR Plus perks at plus.npr.org. Back now to our discussion about how AI is reshaping education. and how teachers and students feel about it. 1A producer Haley Blassingame recently spoke to high school senior David McCann-Lieres of Pickerington High School North in Ohio.

13:15He was one of 98 students from across the country selected to convene at the Edward Kennedy Institute this summer to pass mock legislation to help schools confront the question of AI in schools. Haley began by asking David about his relationship to AI. Right now, I think of it very differently as when consumer AI entered my mind. I thought it was like a fun little toy when I first heard of it in like eighth grade. I'm a senior in high school now. But back when the first version of ChatGPTs were going public, I thought it was a fun toy. I didn't realize it could be used for plagiarism until everyone started going, oh my gosh, it could be used for plagiarism and all that.

13:53These days, I don't really touch it. I don't think it's very conducive to the building of cognitive power and your own ability to think and keep your, and study. I don't use it very much. And what about your peers and your friends? What are they saying about AI and education? I'm in a lot of circles of creatives. So a lot of my friends similarly do not use AI at all, unless they're professors. A lot of my friends are in college by now, the professor says, oh, we have to use AI for this assignment to demonstrate something. Typically, they don't do that. There are some peers of mine who use it for everything.

14:33There's really just a spectrum in every classroom. But on average, I'd say that my friends do not use AI or some of my peers will occasionally try to use it to help complete a tedious task, like brainstorming for an essay. Do you feel like there are any generational differences in how people view AI? Yes. I believe that older generations may see more use of AI in areas which they haven't been more active recently. Like someone who's in their 50s or 60s isn't currently in high school and may not be a high school teacher, but they still think, oh, what if AI could be used in schools? What if AI could be used in this industry?

15:17A lot of the time, I feel like the younger people. When we see AI, we're like, how can I apply this to my life? And the older people look at AI and say, how can I apply this? How can other people apply this to their lives? Or they just ask it simple things like the weather, like you buy an Alexa for in 2015. So you were one of 98 students who gathered in July to try to draft legislation on AI in education. What kinds of questions were you and other students trying to answer at that summit? I think it was less about the questions we were trying to answer and more about the problems we were trying to solve.

15:55The biggest glaring issue is that Congress has not yet passed any legislation on AI. And while there is good things to things being less regulated, there needs to be some sort of ground rules for AI in education. Some states have begun creating those rules. My state, Ohio, has set a baseline AI usage guidelines to the Department of Education. And a lot of the things, I remember my peer Hunter from Ohio, also my town Pickerington, he said, we've missed the mark. A big recurring theme here is we've missed the mark with cell phones in schools. So how can we stop AI from becoming a problem in schools before it does?

16:38A lot of the things was more about how can we solve a problem? How can we use our experiences as students to make the experience of future students better if this bill were to be a real thing? Yeah. And what do you think the adults, you know, teachers, parents, tech companies, congresspeople, what do you think they're missing in this conversation on AI right now? Partially the voice of young people. Almost never in history do you get to hear from people my age. A lot of the time you think of the Vietnam War, the college students protesting. However, you almost never hear voices of high school students and younger students, which is super important, especially right now because we are discussing education.

17:24And I think this is why I'm so glad I got to participate in that my peers there at the AI conference were able to voice our opinions. I feel like there is a lot of truth in what the AI specialists are saying and what the tech companies PR are saying, what the teachers are saying. Everyone has some truth in what they're saying. However, there is large issues caused by the exclusion of some voices, the ones who are at stake, the stakeholders, which are the students, arguably the biggest stakeholders. Yeah. And finally, what would you like to see from those in power on this? action. The number one thing to do is take a step.

18:05You do not know how big the step will be. You do not know which way it will be in. But if you take a step, the next one will appear. We need to start some legislation, but introduce a bill, bring lobbyists in, whatever is needed to take the first step towards regulating AI, towards making AI safer, making it a better place. there's so many different topics that we covered at the AI Summit. Some were concerned about student safety. Some were concerned about plagiarism and academic integrity. Some were concerned about moral issues or environmental issues. And all of those things are important, but none of those things will get done if we go, we have to fix everything right now, and we don't take the first step.

18:50That's high school senior David Makan Linares speaking to 1A producer Haley Blassingame. I want to get your reactions to what we heard there from David, especially his assertion that adults essentially aren't doing enough. Jeff? Well, I think we put this conference together because we scanned the playing field, and what we saw is, to Professor Reich's earlier point, it's kind of chaos out there. A lot of people doing a lot of different things, but there's been no national regulation. There's I got into this because we did miss the boat on social media and cell phones 15 years ago. And while I think AI is going to be the next revolution in education, there's a lot of bad things with AI.

19:32And so I think we need to put some guardrails in place. And so when we convene the kids, their number one policy statement was that, you know, as soon as devices are entered into in school and put in place in front of kids, we need to have AI literacy. Justin, what about you? what do you think about this assertion that the grownups are not moving quickly enough on this? Well, I absolutely celebrate what the students and Jeff pulled together. I think listening to young people is incredibly important. I think one huge challenge with moving forward, which we have to do something, but we actually do not know what to do.

20:10There is not good research evidence yet that tells us what are effective approaches to AI literacy. Is AI literacy an effective approach? What kinds of regulatory or instructional strategies would work? And, you know, I think the student says something along the lines of, well, we just have to take a step forward. But if you look at the history of education technology, an unfortunate fact is that many of our early guesses about how to incorporate new technologies into schools are wrong. Sometimes they're benignly wrong. I bet many of your listeners have been to classrooms where there are smart boards gathering dust in the back corner.

20:47And unless they fell on someone, those smart boards didn't hurt anyone. But they were an enormous waste of time and money that really didn't help students learn at all. But there are other technologies that we've introduced into schools where we have miseducated millions of students. So I started teaching in the early 2000s when the World Wide Web was just introduced. And when I was a young teacher, the way we taught students to sort truth from fiction on the web was to look really closely at websites looking for markers of credibility. We told them not to trust Wikipedia. We told them to trust.org and.gov websites, more than.com websites.

21:21All of that is wrong. For two decades, we taught tens of millions of students provably incorrect ways of searching the web, sorting truth from fiction from the web. And it really wasn't until the very end of the 2010s, 2019, that we started developing more research-based, better strategies, most of which have not replaced the ineffective strategies that appear in schools. And so there are real risks with moving forward. It's very easy in education to have experts sort of congeal around ideas that prove to be wrong and prove to be harmful in developing students' abilities. So we really need to, whatever we're trying, we need to evaluate.

22:04I want to talk a little bit more about the research on AI and student learning. But firstly, how much does what we heard from David there resonate with what you hear from other students in your reporting? Oh, it resonates a tremendous amount. And I will say that students themselves hold a lot of insight and wisdom on how AI is affecting their learning experiences. Justin is 100 % correct. There is not a robust body of research yet to give us a sense of how this technology is impacting students and how it can be used to benefit them and how it shouldn't be used because it could harm them. But one thing I will say that I hear consistently from students, even students who see value in AI and use it as a tool to enhance their learning, one thing they all recognize is that this tool can be misused.

22:56It can completely bypass the learning process. And that's something that I find young people are consistently worried about. They're worried, for lack of a better term, that this technology is going to make them dumber. Well, after analyzing more than 800 academic papers on AI in K-12 education this year, researchers at Stanford found only 20 high-quality studies that examined how AI tools affect students and educators. And so, Jeff, in your work, how are you creating lesson plans and curriculum to teach AI when the research on it remains so far behind? Yeah, I think one of the interesting things we do, and we kind of posit when we're out there, a lot of districts ask us for help with policies on AI, and they want to do it at the school committee level.

23:44And if you know schools, you know that school committee policies typically last for 10 or 20 years. But this technology is moving so quickly, we don't actually recommend that. We recommend that they start at the high school student handbook level and start from a place of values. We value academic integrity. Great. Then a teacher has to disclose if they want the kid to use AI and the kid has to disclose if they are using AI for a project. But underneath it, I think, is this idea that there are real concerns for kids. And this idea of cognitive offloading where the technology does the thinking for kids and our brains atrophy is a real issue.

24:22And so I think, you know, I look at Connecticut, who has taken what I would consider to be the chaos that's out there and try to control the chaos. And they've done a bunch of different pilots with their teachers, different platforms, different tools. And they're trying to take the data to see what they think is best. And I think Professor Reich's point is true. It's going to take a while for us to really know the best practices, the best AI tools. But with that said, I'm not sure we can actually just wait for that because kids' safety is on the line. And I think there's some common sense things we can do as a community to make sure that we protect our kids.

25:02But when we talk about safety, I mean, there's a question about safety in either direction. If you put a pause on this and say, let's get better research into how this affects students, or if you don't train students on AI. So I'm curious to hear how you're accounting for the risks of forging ahead with AI in an educational setting without hard data to back up its efficacy and safety. Well, the number one thing we promote is AI literacy. We think the kids need to know the pros and the cons of AI before they touch the tools. And we think that's the most important thing, right? You don't just hand your kids the keys to the car when they turn 16.

25:39You put them in driver's ed and then you white knuckle in the front seat with them as they learn to drive. And I think AI is similar, right? We need to give them kind of some of the rules of the road, explain to them about the good things that AI can do, but also the bad things, right? Hallucinations, bias, deep fakes, energy concerns, cognitive offloading, you know, the AI companions where some of these tools have helped kids kill themselves. Like there's real problems here. And I think we need to educate our kids about what AI is and what it's not. Let's go to our inbox. Hi, my name is Quincy and I am from Sacramento, California.